The Creativity Algorithm

Humanities last stand against AI

I have a reputation for being a creative game producer. I am proud of the crazy ideas I was able to realize. I would credit tenacity more, but creativity is a big part of it. You can be highly educated, possess immense experience, or be wildly lucky at failing up; if you’re not creative, you need to partner up with someone who is. You probably know that already, being intelligent.

Creativity is something completely different and should be in its own category.

How does it work? What is the difference between creative and non-creative? Is creativity on a spectrum? And, where does it come from? Seems like eureka moments are more common than not. Solutions to problems seem to just pop up. How often do you get a key idea right when you wake up? Or randomly in the shower. It's like creativity comes from deep within. And there’s something magical about it.

Or is it simple? Like a machine learning algorithm.

Compare us to our new cousins, the LLMs. They recall facts like Wikipedia and connect microscopic data points across an ocean of information. Like finding associations between two needles in a haystack. That is what LLMs do best. As amazing as they are, LLMs rely on simple algorithms.

Human creativity is a different animal. It operates in its own class, driven by distinct mechanics.

First, it just happens. No one sits down and says, “It’s time to be creative.” Instead, it works like this. You encounter a problem. The desire to solve it compels you to build a mental simulation of the problem. Once the simulation of the problem is running in your mind, you throw potential solutions at it. The simulation tests the validity of the ideas that come out of nowhere.

In other words, you don’t formulate ideas; they spontaneously occur; you only stress-test them.

The ideas themselves originate from elsewhere. They feel spontaneous, perhaps even random. Broken down to a simple algorithm, the process would look like this:

1. Take two random things. Like a gummy bear and a penny.

2. Put them together.

3. Judge if the combination works. Hmmm.

If it fails, you discard them and roll again. Mathematicians often note that insights arrive completely unprompted; you have to wait for a good idea to show itself. Patience enhances creativity purely by allowing more rolls of the dice.

So what about those instantaneous flashes of insight? Where do they really come from? Do we have quiet voices in our heads, thinking deeply in the background, occasionally breaking the silence to whisper an idea that we could not have consciously constructed alone? Can we say that when we look closely, consciousness itself is not creative?

Perhaps as humans, our primary job isn’t creation but curation. Most random combinations turn out to be not worth a 2nd thought. The trick is to stay patient, keep the simulation running, and pay close attention so you don’t miss the whispers of creativity. Attention is all you need.

This is me thinking about creativity in the context of ideas and artificial intelligence. When I contemplated the above image, I tried to come up with two of the most random objects I could think of. After I came up with the gummy bear and a penny, I put them together in my mind. Took a step back and considered the idea.

Luckily, as I said, I have tenacity, so I can be patient and wait for the next lightning strike.